China Mobile demonstrated 1.6 Tbps signal transmission on a single wavelength over 960 km of standard fiber

The research institutes of China Telecom, YOFC, and ZTE have completed the implementation of what they call the world's first real-time bidirectional data transmission over fiber optic cable spanning over 1000 kilometers, as announced by the China Telecom Research Institute on November 2, according to sources IT Home and C114.

On a line measuring 1007.48 km, the partners achieved a speed of 800 Gbps per wavelength using PCS-16QAM modulation in the extended C+L band, transmitting signals in both directions through a single optical fiber cable, resulting in a total capacity of 128 Tbps. These three organizations collaborated as joint builders of a national key laboratory.

The partners reported three significant achievements. This experiment set a record for the longest single-point, non-loop real-time data transmission over fiber optic cable while maintaining a loss margin of at least 3 dB on each segment and an Optical Signal-to-Noise Ratio (OSNR) margin of at least 3 dB for system operation. Furthermore, it became the first real-time bidirectional transmission over a single optical fiber line at the 1000 km scale.

Although reflections reduced the OSNR by between 0.5 dB and 4 dB, the team believes this proves the feasibility of long-haul bidirectional operations. Finally, the longest fiber optic line in the extended C+L band free from gas line absorption was constructed.

The last point concerns an issue specific to this technology. Since light in hollow-core fiber passes through gas rather than solid glass, traces of carbon dioxide and carbon monoxide in the core can absorb certain wavelengths and degrade transmission. To solve this problem, the team improved the manufacturing process to eliminate these absorption lines across the entire extended C+L band.

Regarding equipment, the partners eschewed high-power amplifiers and used commercial amplifiers with standard output power, matched to the fiber loss profile, to achieve optimal performance from a fully loaded 800G system. Light pre-equalization at the transmitter level maintained channel stability across the entire band. The team also developed an end-to-end model of the impact of reflections and intermodal interference on the line and tested an extreme scenario where part of the band failed without detecting any performance loss.

Operators are most interested in latency. Light in hollow-core fiber travels almost at the speed of a vacuum, approximately 30% faster than in traditional glass fiber, saving about 1.5 milliseconds in each direction over a 1000 km route. The announcement did not specify the measured latency value. Until now, hollow-core fiber in China has mainly been used for short connections, such as between data centers in Chunwei, Ningxia, whereas computing clusters in the west are often more than 1000 km away from demand in the east. China Telecom stated that the result provides a practical technical basis for creating high-capacity, long-distance, and low-latency backbone networks based on fiber optic cable.

A research group led by Chen Mohan from Peking University adapted ABACUS—an open-source density functional theory (DFT) package developed in China—for the full line of Huawei Kunpeng 920 processors. Furthermore, the team optimized 19 different classes of core operators. This was reported by Huawei Computing in a publication on October 31, as learned by IT Home. The code was uploaded to the project repository on GitHub.

Density functional theory (DFT) is a primary first-principles method for studying the properties of materials such as metals and semiconductors. ABACUS was jointly developed by several institutions and has over a thousand users across more than one hundred organizations.

According to the team's data, when configuring both sides to 28 cores with 28 processes running one thread each, the Kunpeng 920 build, utilizing the SVE instruction set, showed speeds 1.3 to 3.9 times faster than a standard x86 CPU when working with a standard set of 10 material systems containing 8 to 216 atoms, including GaAs, MoS2, BaTiO3, iron, and silicon. When using the chip's matrix computation unit, the speedup reached up to 10.7 times when operating on all cores of a single node. In cases of stochastic DFT, the team recorded a speed increase of 2.8 to 3.1 times for SVE and 6.1 to 20.1 times for the build using the matrix block.

The initial step was porting, as various Kunpeng 920 models support vector instructions NEON, SVE, or SVE512, as well as a dedicated matrix block, which remain unused when using general compiler options. The team created separate toolchains: gcc with OpenMPI and OpenBLAS was used for NEON and SVE builds, while clang with HMPI and KBLAS from Huawei was used for others, and the USE_KML option was added for integration with the Kunpeng Math Library. All compiled versions passed ABACUS integration tests covering DFT for plane waves and LCAO Kohn-Sham, stochastic DFT, and DeePKS, with results matching the main implementations.

Next, the team rewrote frequently called small operators. Replacing the recursive procedure for the generalized Laguerre polynomial with an iterative recurrence allowed complexity reduction to O(n) and made it approximately eight times faster in a million-test run. Switching interpolation search to binary search reduced the average number of comparisons in a 400-entry table from about 200 to nine. Other changes included trigonometric recurrences, cycle invariant lifting, mixed-precision tuning, and memory layout, with the numerical results of each operator being re-verified.

Using automatic source code optimization in Huawei's DevKit, the team also increased the processing speed for an 8-atom high-density carbon case by 2.28 times and a 512-atom silicon case by 2.66 times on the newer Kunpeng 920 model. Huawei notes that the resulting tuning experience can be applied to other scientific software.